Missing e mails

[Sociable]

Precisely my point stuartbe

I'm not talking about NTL applying one of the domestic anti-spam programs but instituting a process that identifies the source of spam and applies a block.

Examples would be attempts to send the same mail to a dictionary list clearly designed to test for valid addresses and other such methods used by the spammers.

I'm no tech but simple logic tells me it must be possible to identify and eliminate some of the more obvious abuses of the system and start with those.

No more nor less than that.

[Stuartbe]

Originally Posted by Sociable

Precisely my point stuartbe

I'm not talking about NTL applying one of the domestic anti-spam programs but instituting a process that identifies the source of spam and applies a block.

Examples would be attempts to send the same mail to a dictionary list clearly designed to test for valid addresses and other such methods used by the spammers.

I'm no tech but simple logic tells me it must be possible to identify and eliminate some of the more obvious abuses of the system and start with those.

No more nor less than that.

I see what you are saying m8

The problem remains that there is no 100% efective method of detecting spam. Until we can do this we cant start to block. The risk of killing personal mail is too high.

Microsoft have a duty here to !! they should make sure that the mail clients are able to deal with junk mail. Reverse dns checking is just one method that is very efective.

If we are to rid the internet of spam (something I think we will never do as the internet is based on totaly floored technology) then everyone has to work together. How many users know how to check the source ip (spoofed or not) of an email ? How many users foolishly click on the unsubscribe link ?

We could argue until the cows come home (or the wife) but there is no out of the box solution.

[Sociable]

Fair point.

But saying that as we can't eliminate 100% means we shouldn't eradicate a lower percentage that is known to be safe is another matter.

Start with the clear abuse first and work down the list.

Yes even this could result in a few stray messages being caught by the system but only where people are routing mail through insecure or deliberately spam friendly servers. This in turn would lead to people stopping support for those servers and forcing them to clean up their act.

Curiously I can remember a time when the old cabletel email address was on a blacklist because it was using insecure mail servers. AFAIK it was this being blacklisted that concentrated the minds of the ISP to fix the problem.

[Stuartbe]

Originally Posted by Sociable

Fair point.

But saying that as we can't eliminate 100% means we shouldn't eradicate a lower percentage that is known to be safe is another matter.

Start with the clear abuse first and work down the list.

Yes even this could result in a few stray messages being caught by the system but only where people are routing mail through insecure or deliberately spam friendly servers. This in turn would lead to people stopping support for those servers and forcing them to clean up their act.

Curiously I can remember a time when the old cabletel email address was on a blacklist because it was using insecure mail servers. AFAIK it was this being blacklisted that concentrated the minds of the ISP to fix the problem.

O.K. m8

Lets put it to the poll and see the results !!!

I think that we will have to agree to disagree on this one

Poll - http://forum.nthellworld.co.uk/showthread.php?t=6015

[SMHarman]

Originally Posted by Sociable

Fair point.

But saying that as we can't eliminate 100% means we shouldn't eradicate a lower percentage that is known to be safe is another matter.

Start with the clear abuse first and work down the list.

Yes even this could result in a few stray messages being caught by the system but only where people are routing mail through insecure or deliberately spam friendly servers. This in turn would lead to people stopping support for those servers and forcing them to clean up their act.

Curiously I can remember a time when the old cabletel email address was on a blacklist because it was using insecure mail servers. AFAIK it was this being blacklisted that concentrated the minds of the ISP to fix the problem.

Brightmail operates several spam detection systems, but a key benefit is that they have set up a large number of dummy email addresses at many different ISPs (including Demon). Any email sent to these addresses is by definition unsolicited and unwanted. Brightmail then create filtering rules that will detect the same email being sent to other addresses. This scheme means that Brightmail will not block properly run mailing lists, legitimate mailshots to customers or indeed mass emails sent by your friends to invite you to a party. This is because these emails will never reach the dummy addresses and so no rules will be created for them. For more information see:

http://www.brightmail.com/accuracy.html

[Stuartbe]

Originally Posted by SMHarman

Brightmail operates several spam detection systems, but a key benefit is that they have set up a large number of dummy email addresses at many different ISPs (including Demon). Any email sent to these addresses is by definition unsolicited and unwanted. Brightmail then create filtering rules that will detect the same email being sent to other addresses. This scheme means that Brightmail will not block properly run mailing lists, legitimate mailshots to customers or indeed mass emails sent by your friends to invite you to a party. This is because these emails will never reach the dummy addresses and so no rules will be created for them. For more information see:

http://www.brightmail.com/accuracy.html

You dont work for them do you ?

Seriosly though - Looks pretty good !! I think the price tag may be out of NTL'S bracket though.

I cant get the picture of two monkeys reading email on a screen out of my head

[SMHarman]

Originally Posted by stuartbe

You dont work for them do you ?

Seriosly though - Looks pretty good !! I think the price tag may be out of NTL'S bracket though.

I cant get the picture of two monkeys reading email on a screen out of my head

No I don't, I use Demon for overpriced web hosting and pretty good shopping cart and credit card acceptance. I initially used them for dial up and still for backup ISDN access (free with shopsite) when NTL is pants. I also use theirs not NTLs POP servers (but NTLs SMTP servers).

The Spam issue is interesting as Demons POP servers have been through significant growing pains over recent months as they collapsed under the weight of spam. A big upgrade has solved this, but it looks like they want to capacity manage (and IMHO improve the end user experience) by implementing filtering.

Being as my mail box got 600 spam mails between 23/12 and 4/1 I will be able to report back on the results.

[Sociable]

Great post SMHarmon

A very good example of exactly the sort of thing I mean.

[threadbare]

Originally Posted by stuartbe

Sorry m8 - its very likely.... Spam asasin for example will spam mark a mail that does not have a subject header, How many time have you sent or been sent an email without a subject ????

i believe ntl mail filters use an IP blacklist not subject headers

[Stuartbe]

Originally Posted by threadbare

i believe ntl mail filters use an IP blacklist not subject headers

So are you saying that an IP blacklist cant block mail that is genuine ???

I use ordb, reverse dns check, header filters and spam asasin and I have had mails from my bank blocked by ip blacklists. All it takes is for a spammer to spoof the ip of there mail server and they get blocked. O.K. it can be unblocked, but not before mails get zapped.

I use all the spam detection methods above but I still only mark the mail as spam and sent it to a diferent folder. I have had multiple emails that have ended up in the spam folder that were perfectly o.k.

I am not for a moment saying that NTL should do nothing. I am saying that you need to be 100% sure that a mail is spam before you block it, and then only with the consent of the customer. Emails are private to the user and can sometimes be personal to. Some emails are even treated with the same legality as postal letters if they have digital sigs.

I wonder how many people would rant and rave on this site if they had a personal mail blocked ?

[threadbare]

Originally Posted by stuartbe

So are you saying that an IP blacklist cant block mail that is genuine ???

no i didn't say that at all. please see my previous post as to what i said.

edit: perhaps a tech can come on and explain exactly what ntl's mail filtering policy is before we get ourselves into a little frenzy

(we as in nthw not you in particular stuartbe)

[threadbare]

I have noticed a drop in the amount of spam i get since the filtering was introduced. which was in december i believe

Frankly I just want to get less Viagra spam,diet pills offers,job offers,cheap tobacco offers,mortgage and finance offers I'm not interested in.It's as bad as all them bluddy leaflets in the magazines that litter everywhere.

At least the teenage girls sex with horses spam emails has stopped about 3 months ago which was a great improvement but I wonder WHY it stopped as I did nothing about it.I found those particularly annoying.

As for NTL spam filtering could they at least give it a go?If it doesn't work they could always think again surely?Something is going to have to be done soon before the whole system grinds to a halt.

Incog.

[greencreeper]

Originally Posted by Sociable

Second if mail is refused the sender would be adivised his mail was rejected and why so appeals can be made.

Not true for mailing lists. You'd would be taken off the list. You would complain and re-subscribe. You would be taken off the list. You would complain and re-subscribe. You would be... In today's world when the sender is often a machine you cannot expect the machine to sort out the problem OR make it's owner aware there is a problem.

The trouble with blacklists is that a lot of them are open to the (muppet) public to add "spammers" and the (muppet) public often adds legitimate senders.

There's also a problem when a company uses the same mail server/email address to send unsolicited emails (that some would see as spam) and legitimate emails to registered customers. Customers end up losing emails or receiving them marked as spam.

Many customers sign up to a site and don't tick the "I don't want a newsletter" box. When they get the newsletter they promptly run off to the nearest blacklist database and report the sender.


john

[paulyoung666]

from my point of view though ppl should have to opt in , not opt out